Police Training Is Seriously Lacking in Actual Science

Getting new ideas from social science into police academy training and thousands of local police departments is a slow endeavor.
Demonstrators marking the oneyear anniversary of the shooting of Michael Brown face off with police during a protest...
FERGUSON, MO - AUGUST 09: Demonstrators, marking the one-year anniversary of the shooting of Michael Brown, face off with police during a protest along West Florrisant Street on August 9, 2015 in Ferguson, Missouri. There are reports that two people were shot when gun fire broke out during protests later in the evening. Brown was shot and killed by a Ferguson police officer on August 9, 2014. His death sparked months of sometimes violent protests in Ferguson and drew nationwide focus on police treatment of black suspects. (Photo by Scott Olson/Getty Images)Scott Olson/Getty Images

Michael Brown was, at best, stopped by police for stealing cigarillos. Sandra Bland for failing to signal a lane change. Freddie Gray for carrying a switchblade. Yet these encounters all ended with them dead. Distrust running both ways between police and the communities they’re supposed to protect have sparked cries for reform to prevent rapid escalation of police violence. What’s missing in the conversation, though, is science.

That’s because the science often doesn’t exist. Police rarely cooperate with outside researchers, especially those perceived as reformers. “In New York where I’ve done a lot of my work, I can’t get anyone to talk to me,” says Alex Vitale, a sociologist at Brooklyn College who has studied how police respond to protests. And even when social science research points to a need for reform, getting new ideas into police academy training and thousands of local police departments fractured all over the country is, put charitably, a slow endeavor.

Since 1994, the Department of Justice has funneled more than $14 billion to state and local police departments to support community policing initiatives, to limited success. The idea behind the so-called Community Oriented Policing Services program comes out of contact theory, which suggests the best way to reduce prejudice is to interact with people who are different. Police officers—particularly white officers working in black neighborhoods—only encounter people different from them when they’re responding to a crime. They rarely see the community in a positive light. Get officers out in the community in ordinary situations, and they’ll become less defensive and negative. That much bears out in the science.

The big question, though, is what that means when officers actually hit the streets. More foot patrols? Knocking on doors to chat? “We’re still trying to figure that out,” says Charlotte Gill, a criminologist at George Mason University. Gill looked at 65 community policing studies for a 2014 meta-analysis she co-authored, only to find no clear evidence that the approach reduces crime. The results are so muddled, in part, because there is no concrete definition of community policing. The US has spent $14 billion on an undefined goal.

That’s especially frustrating because the science that exists about bias—police or otherwise—clearly points to a need for reform. Dozens of psychology studies show that people hold implicit biases against African-Americans—they might not admit or even be aware of these biases, but reaction times give them away. Police statistics bear them out in the real world, too: The Department of Justice’s investigation into Ferguson found that police were twice as likely to search blacks than whites during traffic stops, even though contraband was found 26 percent less often with blacks.

And those biases become especially serious when they extend to deadly force. A 2007 video game simulation study of Denver police officers found that that officers are faster to shoot blacks than whites. At least, the authors pointed out, the officers were less likely to mistakenly shoot unarmed blacks than the untrained general public—a result that they chalked up to high-quality gun training for officer, also called use of force training.

But what kind of training really works—and whether it can be designed to reduce the influence of racial biases—is yet another open question. The Force Institute is one of the leading organizations that provides use of force training. Its founder, Bill Lewinski, a former Minnesota State professor, was recently the subject of a profile in the New York Times, which criticized him as a psychologist for hire who got cops off in deadly shootings. (The institute declined a request for an interview with WIRED.)

Little research into this area exists, and what does exist is carried out by people with a vested interest. Lewinski, for example, has argued that suspects can draw a gun more quickly than an officer can draw from a holster and aim, so police are justified in reacting before they see a gun. An American Journal of Psychology editor who reviewed one of the studies for the Justice Department called the research “invalid and unreliable.”

As protests have erupted around some of those preventable police shootings, departments have turned to de-escalation training for officers. Here too, peer-reviewed research is lacking, and even police trainers turn to personal experience. “Most of the stuff we have done and I’ve done is at a personal level. A lot of them use our own experience,” says Gary Klugiewicz, a former sheriff’s department captain who now teaches conflict management at Vistelar. When asked about research into the effectiveness of de-escalation tactics, he pointed to the Force Science Institute’s Lewinski as a leading researcher.

The personal experiences of police officers are of course relevant to effective policing, but police departments have generally resisted lessons from research. Carl Bell, a psychologist at the University of Illinois at Chicago who has done key work on de-escalation with the mentally ill, said his attempts to introduce techniques to the Chicago police never got anywhere. “There’s no systematic incorporation of research,” says Gill, who has a student studying how community policing is being taught in police academies.

Even with training, officers may not think it’s important to use the de-escalation techniques taught. “The universal greeting is taught to most of the country yet it’s not done,” says Klugiewicz, referring to how officers introduce themselves and explain their presence. “Officers get lax. Officers may not think it’s that important. They might be angry.” In the much circulated dashcam video of Bland’s arrest, the officer snaps after Bland refuses to put out a cigarette. He yells her to get out of the car. “You ask ‘Can you get out of the car?,’ not say ‘Get out of the car,’” says Klugiewicz.

But police trainers and scientists alike seem to agree that the moment for change is now. Klugiewicz calls the recent events a perfect storm: “A perfect storm can be really disruptive but it can also create change.” Gill says a police department recently approached her to do a study, rather than the other way around. “It was really surprising to me,” she say. “That rarely happens.”

Science alone obviously doesn’t hold all the answer for policing—there are cultural and social factors to consider. But if we want empirical evidence behind the reforms we demand, then researchers can certainly do better work when they aren’t sidelined in their ivory tower.